The present invention relates generally to backup digital data of storage subsystem and, more particularly, to block-based volume backup involving block storage subsystem such as SAN (Storage Area Network), using file storage subsystem such as NAS (Network Attached Storage), HTTP (Hyper Text Transfer Protocol), and the like for backup target.
Backing up data is very important to prevent data loss from device failure, disaster, operation failure, and so on. Traditionally, a storage subsystem has utilized tape device (D2T), block disk device (D2D), and a combination thereof (D2D2T) for backup target. Today, there is another backup target which is provided by cloud storage services. Generally, a cloud storage service consists of file storage service such as NAS, HTTP, and so on. It is known as D2CD (cloud) backup.
Most of the current solutions are file/file-system based cloud storage backup. For instance, a software program on the host computer reads data via a file-system such as ext3 and sends the data (as a file) to cloud storage services. The data itself is stored on local disk, DAS (Direct Attached Storage), SAN, and so on. However, several application programs such as database require block based backup method, while these solutions allows only file based backup.
Other current solutions are sending block SCSI command to the HTTP server by translating host SCSI read/write command to HTTP GET/POST command. It allows representing the HTTP server disk space as volume, LU (Logical Unit) on the storage subsystem. However it means direct command conversion and it lacks the method of creating multi-generation snapshots for backup. A backup system always requires multi-generation backup images, so that the methods of creating multi-generation snapshot must be required.
Currently, there is a volume-based backup method by utilizing a storage subsystem, but it can be carried out only with block base storage subsystems.